Algorithm 1. Weighted voting scheme.

Input:

${\mathrm{C}}_{\mathrm{i}}$: Classifier

${\mathrm{L}}_{\mathrm{j}}$: Labels of Data Set

m: Ensemble Size

n: the Number of Labels

Output:

the predicted class ${\mathrm{y}}_{\mathrm{j}}$ from a single classifier ${\mathrm{C}}_{\mathrm{i}}$

the predicted class y*

for i = 1: m

for j = 1: n

compute ${\mathrm{p}}_{{\mathrm{C}}_{\mathrm{i}}{\mathrm{L}}_{\mathrm{j}}}$, the probability assigned by ${\mathrm{C}}_{\mathrm{i}}$ to class ${\mathrm{L}}_{\mathrm{j}}$

$\mathsf{\mu }$ = ${\arg \max }_{\mathrm{L}=1,\dots ,\mathrm{r}}$ ${\mathrm{p}}_{{\mathrm{C}}_{\mathrm{i}}{\mathrm{L}}_{\mathrm{j}}}$

${\mathrm{y}}_{{\mathrm{C}}_{\mathrm{i}}}$ = ${\mathrm{y}}_{\mathsf{\mu }}$

for j = 1: n

$\mathrm{y}\left({\mathrm{L}}_{\mathrm{j}}\right)$ = { i = 1,…,m: ${\mathrm{y}}_{{\mathrm{C}}_{\mathrm{i}}}$ == ${\mathrm{y}}_{{\mathrm{L}}_{\mathrm{j}}}$}

if $\mathrm{y}\left({\mathrm{L}}_{\mathrm{j}}\right)$ == $\varnothing$

${\mathrm{g}}_{{\mathrm{L}}_{\mathrm{j}}}$ = 0

else

for i in $\mathrm{y}\left({\mathrm{L}}_{\mathrm{j}}\right)$ do

${\mathsf{\omega }}_{{\mathrm{C}}_{\mathrm{i}}}$ = ${\max }_{\mathrm{j}=1,\dots ,\mathrm{n}}$ ${\mathrm{p}}_{{\mathrm{C}}_{\mathrm{i}}{\mathrm{L}}_{\mathrm{j}}}$

${\mathrm{g}}_{{\mathrm{L}}_{\mathrm{j}}}$= ${\displaystyle \sum }_{\mathrm{i}=1}^{\mathrm{m}}{\mathsf{\omega }}_{{\mathrm{C}}_{\mathrm{i}}}$,

$\mathsf{\mu }$ = ${\arg \max }_{\mathrm{L}=1,\dots ,\mathrm{r}}$ ${\mathrm{g}}_{{\mathrm{L}}_{\mathrm{j}}}$

y* = ${\mathrm{y}}_{\mathsf{\mu }}$

return y*